APRIL
1958
589
PSORALENES I
TABLE I (Cont'd) SUBSTTTUTED HIPPURICANILIDESFROM HIPPURICACID A N D SUBSTITUTED ANILIXES tion in Hours 138-162
.
hr.
162-380
hr .
Solvent for Recrystallization
Nitrogen Analyses yo Calcd. % Found
Total Yield, Grams
Percentage Yield
Color of Product
0.0982
0.4076
Ethanol
15.51
See Ref. 4
4.9539
36.75
White
0.0850
0.5430
Ethanol
15.51
15.56
6,1500
45.62
White
0.161
0.725
Ethanol
10.37
10.45
3,7900
28.00
White
0.0026
0.0000
Ethanol or ethanol
10.37
10.2G
6. '3844
51.69
Cream
0.177
0.653
Ethanol
10.37
10.38
9.4690
i 0 09
White
0.39
9.29
5.4380
36,47
Cream
9.30
0.32
1,9330
14.34
White
0.46
0.40
10.8702
'73.35
Cream
acid, and p-nitroaniline' gave no precipitate 0.0791 0.3305 Glacial HOAc; then addition of water 0.0120 0.153 Glacial HOAC; then addition of water 0.091 0,3162 Ethanol or acetophenone 0.2111
1.1169
Ethanol
9.4G
9.34
3.7810
25.51
Khite
0.2266
1.161 1
Ethanol
9.86
9.87
5.2347
36.81
Cream
0.2437
1 ,0495
Ethanol
9.8G
9.99
11.2409
79.05
White
0.0941
0.3942
Ethanol
8.97
8.96
1.7703
11.71
Khite
[CONTRIBUTION FROM
THE 1)EPARTMENT O F
ClIEhlISTRY, OREGON STATE COLLEGE]
Psoralene I : Certain Reactions of Xanthotoxin* MERVIN E . BROKKE
ASD
BERT E. CHRISTEKSEN
Received May 6 , 1967
-4number of the chemical reactions of 9-methoxypsoraleiic and its derivatives are described. These include nitration, halogenation, reduction, thionation, demethylation, ozonation, and other degradation procedures. Degradation studies and unequivocal synthesis show that bromination occurs at the four position. The structures of the 2,3-dihydropsoralene derivatives were established by a comparison of the ultraviolet absorption spectra of psoralene and coumarin derivatives. Thionation of the 9-methoxypsoralenes proceeds in a manner analogous to that reported for coumarins.
In 1911, Priess' discovered a new piscicide in an alcoholic extract of fagara zanthoxyloide,s Lam. to which he gave the name zanthotoxin. Later Thoms2 after determining the structure of this compound renamed it xa,uthotoxin and in 1036 Spath reproduced it ~ynthetically.~ In addition to its toxic action OH fish4xanthotoxin (n-Inethoxypsoralt.ne or 0-methoxyfuro[3,2-g]coumarin I) has since been shown to possess a mollusci-
cidal a c t i v i t ~When .~ administered in large doses to mainrnals it was found to produce fatty degenerat ion of the liver and adrenal hemorrhagr,Rwhile in humans the compound has found medical acc*cptancefor the treatment of leukoderniia.' The most recent applications have made use of the fact, that, I alters the erythermal response to ultraviolet lightJ,sa*b~c a property which has beeii used clinically to prevent sunburn.** Therc is some cvitlcncc that
* This work was supported in part by grants from the Division of Research Grants and Fellowships, National Institutes of Health, Public Health Service. Published with the approval of the Monographs Publications Committee, Oregon State College as Research Paper KO.327, School of Science, Department of Chemistry. (1) H. Priess, Ber. Pharm. Ges., 21, 227 (1911). (2) H. Thoms, Ber., 44, 3325 (1911). (3) E. Spath and M. Pailer, Ber., 69, 767 (1936). (4) E. Spath and F. Kuffner, Monatsh., 69, 75 (1936).
( 5 ) A. Schiinberg and N. Lntif, J . Am. C'hem. SOC.,7 6 , 6208 (1954). (6) A. Elwi, J . Roy. Egypt. Med. Assoc., 33, '773 (1950). ( i ) I. Fahmy and H. Abu-Shady, Quart. J. Pharm. and Pharmacol.,21,499 (1948). (8) ( a ) A. Lerner, J . Invest. Dermatol., 20, 299 (1953). ( b ) A. Griffin, M. O'Neal, and T. Fitzpatrick, Congress of intern. biochem., Brussels 1955, 121. (c) L. Musajo, G. Rotighiero and G. Caporale, Chemica e industria, 35, 13 (1953).
590
BROKKE AKD CHRISTENSEN
VOL.
23
under certain conditions xaiithotoxiii may be carcinogenic.8b Because of the wide-spread and increasing interest in xanthotoxin for its pharmacological action, this study was undertaken to investigate some of the chemical properties of the compound and to prepare new derivatives for biological testing. Both Priessl and Thoms2 prepared mononitro derivatives of I. Thoms and Baetckeg established that this nitro substituent was at the 4-position by reducing it to the corresponding amino compound, followed by oxidation to the quinone. This quinone was shown to be identical to that obtained from bergaptene (4-methoxypsoralene). I n the current study both the nitro and amino derivatives (see I1 and 111, Fig. 1) were prepared in excellent yields. 4-hmino-9-methoxypsoralene I11 exhibited the nornial behavior of an aromatic amine, for example, in the formation of an acetyl derivative and a Schiff's base. Since I11 appeared to be a potential key conipound for the preparation of other derivatives, it was subjected to a number of diazotization reactions. This type of reaction finds precedent in the work of Dt.y and Kutti1" who diazotized and coupled a group of substituted couniariris; and in that of Noguchi arid Kawanami" who prepared 9-hyt l r o s y - ~ - m e t h o ~ y p s o r ~ ~from l e i ~ eaniinobergaptene. Analogous SandIneyer reactions with I11 yielded the corresponding bromo and chloro derivatives while hydrolysis of the diazotized product yielded the phenol, 4-hydroxy-9-niethoxypsoralene IV. Spath has reported a phenol12 which was shown or its to be either 4-hydroxy-9-methoxypsoralene isomer !)-hydroxy-4-methoxypsoralene. 9-Hydroxy-
4-methoxypsoralene was shown to have a melting point of 198",l1 while I V prepared in this laboratory melted at 220-226". Since Spath found the melting point of his phenol to be 224-226" his compound was probably IV. Schonberg and Sinal3 deniethylated 9-methoxypsoralene to form the corresponding phenol, by the use of magnesium iodide and sulfuric acid. Later, Schonberg and Ayiz14 reported that aniline hydrochloride was a superior cleaving agent. The latter work however, could not be confirmed in this laboratory. Cleavage with aniline hydrochloride was attempted both with the refluxing technique of Schonberg and by fusion in sealed tubes at various temperatures and heating times. In no case could any reaction be observed. Cleavage with the magnesium iodide procedure, on the other hand, was accomplished in small yield. Although other workers have reported cleavage of a butylfurocoumarin ether by mild treatment with mineral acid,16 this reagent as expected was ineffectual with 9-methoxypsoralene. In the original synthesis of I, Spath3 prepared 2,3-dihydro-9-niethoxypsoraleneVI11 which was subsequently dehydrogenated. In this laboratlory it was found that the reverse reaction readily occurred; VI11 was formed in good yield in palladiumcatalyzed hydrogenation react ions. That the hydrogenation product of 9-methoxypsoralene and the 2,3-dihydro-l)-methoxypsoralene reported by Spath were identical structures was deduced from a consideration of the fact that their melting points were identical and from a study of the ultraviolet data given in Table I.
(9) H. Thoms and E. Baetcke, Ber., 45, 3705 (1912). ( I O ) B. Dey and V. Kutti, Proc. Nat. Inst. Sci. Indza, 6, 641 (1940). (11) T. Noguchi :tnd 31. hamanami, Ber., 71, 1428 (1938). (12) E. Spath, Noitatsh, 72, 179 (1938).
(13) A. Schonberg and A. Sina, J . Alnz. Chem. Soc., 72,4826-5 (1950). (14) A. Schonberg and G. A? iz, J . Am Chem. Soc., 75, 3265-6 (1953). (15) G. Pigulevskii and G. Kusnetrovx, Zhur. obshchei Khim., 23,1937 (1953).
APRIL
1958
ULTRAVIOLET
Compound
L"
A B C D E F G H
3.6 4.13 3.28 4.30 4.24
Ll,X,
Log e
...
...
270 255 250 275
3.30 3.70 3.48 3.45 3.52 3.96 3.44 4.29 3.67
283 265
262 246 258
Coumarin16 3,4-Dihydrocoumariii16 6,7-Dihydroxy~oumarin~~ 3,4Dihydr0-6,7-dihydroxycoumarin~~ Benzofuran17
Log
4.04 3 30 3.76 3.60 3.48 3.48 3.90
275 275 295 290 282 289 289 284 290 295
3.40
3,98 3.8s
F G H I J
Log e
e
315
...
350
...
...
...
300 335 >320 >320
3.76
...
4.07 ...
...
... 4.04 4.08 >3.90, >4.21
2,3-Dihydrobensofurani7 9-Methoxypsoralene 2,3-Dihydro-9-methoxypsoralerie 4,5-DimethylpsoraleneC 2j3-Dihydro-4,S-dimethylpsoralene'
This compound has a split peak a t 243 and 248 each with same extinction coefficient. It also has a peak bel017 220. 'This compound has a peak below 220. These two compounds were synthesized in this laboratory and will be published elsewhere.
It is evident from the data presented in Table I that reduction of coumarin and 6,7-dihydroxycoumarin resulted in a lower extension of the conjugation with elimination of peaks above 300 mp.; this same observation has been made in numerous other instances. l6 Furthermore since neither benzofuran, nor 2-3-dihydrobenzofuran shows an absorption spectrum above 300 mp, it is reasonable to assume the benzofuran portion of the furocoumarin molecules would not be responsible for absorption in this region. Therefore the peaks observed above 300 mp, in the cases of G, H, I, and J must arise from the conjugation of the lactone carbonyl with the aromatic nucleus. For this reason the hydrogenation of 9-methoxypsoralene must have occurred in the 2,3-positions rather than the 5,6. It is to be noted that hydrogenation of the furan ring results in formation of a saturated ether which causes a bathochromic effect, offsetting the effect of the loss of conjugation due to reduction of the furan ring. Nitration of 2,3-dihydro-9-methoxypsoraleiie VI11 proceeded simi1:trly to that of 9-methoxypsoralene yielding a niononitro derivative. This nitro derivative resisted dehydrogenation. Although this compound was hydrogenated further by both catalytic. atid chemical reduction no pure products were isolated froin the reaction mixture. On the other hand, 9-methoxy-4-nitropsoraleiie I1 was readily reduced by catalytic procedures to the corresponding 4-amin0-2~3-dihydro-9-methoxypsoralene (IX). The position of the additional hydrogen atoms (2,3) was ascertained by converting the 4-amino substituent to the 4-bromo analog X and then comparing ultraviolet spectra of X with that of the known 4-bromo-9-methoxypsoralene. See Table 11. (16) R. Goodwin and B. Pollock, drch. B~ochenz.,49, 1 (1954). (17) J . Joneb and -1.Lindsey, J . Cheni. SOC.,1836 (1950).
The presence of the peak at 330 mp in the spectrum of 4-bromo-2,3-dihydro-9-methoxypsoralene is indicative of unsaturation in the 5,6 position thus extending conjugation of the carbonyl moiety to the aromatic nucleus. Despite the loss of conjugation due to reduction of the furan ring, the effect of the formation of the aliphatic ether causes a slight bathochromic shift in this instance. Priess18 reported that direct bromination of 9methoxypsoralene in chloroform yielded a dibromo addition product which melted a t 164". He postulated that the addition took place either in the 2,3or 5,6-positions. In this laboratory, the bromination in chloroform yielded a monobromo derivative YII, m.p. 185-186", and a tribromo derivative XVII, m.p. 165". Bromination with N-bromosuccinimide yielded the same monobromo substitution product VI1 as before. Under the various conditions of brominations which mere tried, these were the only products ever isolated from the reaction. The tribromo derivative XVII, was readily converted to the monobronio derivative by the conventional method (treating it with an acetone solution of sodium iodide). Horning and RisnerIg have reported the bromination of 5-1~iethyl-2,~-dihydrofuro[3,2-~]coumarin with M-hrornosuccinimide. Since their product could not he dehydrohalogeriated and did not react with silver nit rate, these workers convluded that the bromine must be either on the lactone ring or the benzene nucleus. The monobronio derivative of 9-niethoxypsoraleneVII was inert to aqueous base, silver nitrate, and Grignard formation under all conditions tested. It was howex-er very labile to catalytic hydrogenation
(18) H. Priess, Chem. Zentralblatt 11, 04 (1911). (19) E. Horning, and D. Rimer, J . A T I L('hem. . SOC.,72, 1.514(1950).
592
VOL.
23
Figure 2
yielding VI11 in 15 minutes a t 40 lbs. hydrogen pressure in presence of palladized charcoal. I n order to establish the position of the bromo substituent, the monobromo derivative was oxidized R ith hydrogen peroxide20yielding furan-2,3dicarbovylic acid, thus eliminating from consideration the 2- and 3-positions. Comparative ozonizations mere carried out using 9-methoxypsoralene and bromo-9-methoxypsoralene, the former yielding the known 6-formyl-7-hydroxy-8-methoxycoumarin. 2 1 Final confirmation of this product was made by the preparation of the dinitrophenyl hydrazone derivative. Bromo-9-methoxypsoralene on the other hand was c1e:tved in both hetero-rings yielding a substituted isophthalaldehyde (XI). The presence of two formyl substituents in XI, was established by the preparation and analysis of the bisphenylhydrazone derivative. AI thoiigli only OR? hetero ring in 9-methoxypsoralene was cleaved by direcl ozonolysis, the corresponding isophthalsldehyde n-as obtained by opcnL ing the lactoiic ring with dimethyl sillfat e . 2 ~ 2IJpori ozonolvsis of the rrsiiltaiit protlurt thr t.:tiii~ reactions ivver(' ui~scrvrd :ts in the instance uf tlic i,l.omo-I)-metlro~yr)~or:llciic. Tlicse reactions :ill pointed to the f:ict that the iiiono1)ron~o suhstituent mud, he in the 4- pusitioii. This observation \vas confiniied \\ hen the 4-bromo-deri\.ative w : ~ i synthesized unequivocally by the Sandmeyer procedure from 4-amino-9-methoxypsorslene. These series of reactions are given in Fig. 2.
The position of the remaining two bromo substituents in the tribromo-9-methoxypsoralene was established by the ultraviolet spectral data shown in Table 11. TABLE I1 ULTRAVIOLET A B s o R P r I o N DATAOF BROMINATED PSORALENE
222
